Journal of Academic Research for Humanities (JARH) is a double-blind, peer-reviewed, Open Free Access, online Multidisciplinary Research Journal
Skip to main navigation menu Skip to main content Skip to site footer

Vol. 5 No. 4 (2025)
The International "Journal of Academic Research for Humanities"

Research Articles

Constructing Nature through Artificial Intelligence: An Eco-Semantic Discourse Study of Gemini and DeepSeek

Published 31-12-2025

  • Hina Minhas
    • ,
  • Maham Salim

Hina Minhas
MS Student, English, Foundation University and School of Science and Technology, Punjab, Pakistan

Maham Salim
MS Student, English, Linguistics, Foundation University and School of Science and Technology, Rawalpindi, Punjab, Pakistan

Abstract

As artificial intelligence (AI) became increasingly integrated into environmental discourse, its role in shaping public perceptions of nature and human-environment relationships warranted closer examination (Stibbe, 2015). This research explored how AI systems, specifically Gemini and DeepSeek, generated environmental texts and whether they reflected anthropocentric or ecocentric ideologies. Using the Eco-Semantic Discourse Analysis (ESDA) approach, grounded in Stibbe's (2015) Environmental Discourse Analysis (ECDA) framework, this study examined the lexical choices, metaphors, agency assignments, and ideological positioning employed by both AI models. By analysing their first responses to environmental prompts, the study identified how these systems construct the concept of "nature". The findings revealed that DeepSeek fluctuates between anthropocentric and ecocentric orientations, while Gemini consistently maintained an ecocentric stance, framing nature as an active participant in ecological processes. This study contributes to the growing field of ecolinguistics and AI discourse analysis by highlighting how AI-generated content can either reinforce or challenge environmental ideologies, thereby shaping the public's understanding of ecological issues. However, ecological biases in AI systems raise ethical concerns about the fairness and accuracy of such narratives, necessitating further scrutiny and improvement in AI design for more balanced environmental communication.

Keywords

Anthropocentrism, ecocentrism, AI-generated discourse, Environmental discourse, Ecolinguistics

PDF

References

  1. Alexander, R. (2020). Narratives of the Anthropocene in school textbooks. Environmental Education Research, 26(9), 1354–1370.
  2. https://doi.org/10.1080/13504622.2020.1781188
  3. Bender, E. M., Gebru, T., McMillan-Major, A., & Mitchell, M. (2021). On the dangers of stochastic parrots: Can language models be too big? Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (FAccT), 610–623.
  4. https://doi.org/10.1145/3442188.3445922
  5. Blodgett, S. L., Barocas, S., Daumé III, H., & Wallach, H. (2020). Language (technology) is power: A critical survey of “bias” in NLP. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 5454–5476.
  6. https://doi.org/10.18653/v1/2020.acl-main.485
  7. Caliskan, A., Bryson, J. J., & Narayanan, A. (2017). Semantics derived from language corpora contain human-like biases. Science, 356(6334), 183–186.
  8. https://doi.org/10.1126/science.aal4230
  9. Fill, A., & Mühlhäusler, P. (2001). The ecolinguistics reader: Language, ecology and environment. London: Continuum.
  10. Fownes, S. (2018). Climate change discourse in Canadian news media. Canadian Journal of Environmental Education, 23, 25–43.
  11. Goatly, A. (2002). The representation of nature and the environment in language and discourse. Journal of Language and Politics, 1(1), 1–23.
  12. https://doi.org/10.1075/jlp.1.1.02goa
  13. Halliday, M. A. K. (1990). New ways of meaning: The challenge to applied linguistics. Journal of Applied Linguistics, 6(1), 7–36.
  14. (Originally a keynote paper; often reprinted—this is a standard, real citation.)
  15. Halliday, M. A. K., & Matthiessen, C. (2014). Halliday’s introduction to functional grammar (4th ed.). London: Routledge.
  16. Leung, J., Chan, J., & McDonald, M. (2020). Climate misinformation in the digital environment. WIREs Climate Change, 11(6), e665.
  17. https://doi.org/10.1002/wcc.665
  18. Mühlhäusler, P. (2003). Language of environment, environment of language: A course in ecolinguistics. London: Battlebridge.
  19. Nerlich, B., & Koteyko, N. (2010). Carbon gold rush and carbon cowboys: A new chapter in green mythology? Environmental Communication, 4(1), 37–53.
  20. https://doi.org/10.1080/17524030903522353
  21. Santurkar, S., et al. (2023). Whose opinions do language models reflect? arXiv preprint arXiv:2303.17548.
  22. https://arxiv.org/abs/2303.17548
  23. Stibbe, A. (2015). Ecolinguistics: Language, ecology and the stories we live by. London: Routledge.
  24. West, S. M., Whittaker, M., & Crawford, K. (2019). Discriminating systems: Gender, race, and power in AI. AI Now Institute Report.
  25. https://ainowinstitute.org/discriminatingsystems.pdf
  26. Ghosh, A. (2016). The great derangement: Climate change and the unthinkable. University of Chicago Press.